KPMG has retracted a major report on artificial intelligence acquisitions after an investigation found it contained fictitious case studies attributed to UBS, NHS Greater Manchester, Transport for London and Swiss Federal Railways – errors identified as AI inventions by research group GPTZero and confirmed by the Financial Times.
The October report, titled “Redefining beauty in the age of agent AI”, said UBS is integrating AI agents across investment advice, risk management and compliance monitoring through a platform developed in partnership with Microsoft. A spokesperson for the Swiss bank told the FT that the claim was untrue.
The report made similar false claims about Swiss Federal Railways, saying the transport group uses AI agents to plan and optimize passenger journeys – a feature the railway confirmed was false. Transport for London said a claim it was using AI agents to predict and manage congestion was misleading, while NHS Greater Manchester said claims about AI-driven patient assessments and hospital readmission predictions were inconsistent with a media release cited as its source.
KPMG International has removed the report from its websites while investigating the circumstances surrounding its publication. A spokesperson said the company expects all employees to follow guidelines on the responsible use of AI, which includes human checks to verify content and verify independent sources.
The episode is not classified. EY withdrew the audit last month after GPTZero identified false footings and other errors, and Sullivan & Cromwell admitted in April that bankruptcy court filings contained AI-generated inaccuracies including a misreading of the US bankruptcy code.
GPTZero, an AI discovery research team that identified errors and whose findings were confirmed by the Financial Times, documented a growing pattern of forgotten content across professional services publications. The credibility attached to firms such as KPMG and EY means that the findings are widely circulated before corrections are issued – the KPMG report was already cited by many industry publications and a major Czech newspaper before it was retracted, increasing the management risk for any organization making its findings.
For CFOs and CFOs who rely on the thought leadership of the Big Four to scale AI adoption or develop internal governance frameworks, the implications are quickly emerging. Research that supports strategic technical decisions at the board level appears to be flawed when produced in volume without adequate human review.
Firms that have cited KPMG’s findings in their AI strategy documents are tasked with validating those references. More broadly, the episode strengthens the case for financial leaders to seek primary source verification before acting on outsourced AI research — regardless of the reputation of the company that produced it.
